Annex to the consultation document - Public consultation...
Transcript of Annex to the consultation document - Public consultation...
Commission européenne/Europese Commissie, 1049 Bruxelles/Brussel, BELGIQUE/BELGIË - Tel. +32 22991111 http://ec.europa.eu/finance/index_en.htm
EUROPEAN COMMISSION Directorate-General for Financial Stability, Financial Services and Capital Markets Union RESOURCES AND COMMUNICATION Internal and external communication
APPENDIX
ECONOMIC ANALYSIS ACCOMPANYING THE CONSULTATION PAPER
ON COVERED BONDS IN THE EUROPEAN UNION
June 2015
2
Table of Contents
I. General description and market trends ..................................................................... 1
II. Covered bonds as an asset class: brief comparison to other jurisdictions .............. 8
III. Potential Risks and Vulnerabilities ..................................................................... 11
IV. Summary ............................................................................................................. 18
V. Schedules .............................................................................................................. 19
3
I. General description and market trends
Covered bonds (CBs) are secured on-balance-sheet debt instruments which give a
preferential claim to the covered bond investors against a ring-fenced pool of collateral
(cover pool) and the proceeds arising from the collateral. Although there is no formal
widely accepted definition for this type of instrument, its main characteristics are:
o the covered bond and the underlying asset pool remain on the issuer's balance
sheet;
o the issuer must ensure that the value of the cover pool dynamically backs the
financial obligations generated by the covered bond;
o the investor has recourse to both the cover pool and the issuer in the event of
default, (dual recourse).
Covered bonds are perceived as low-risk yield-bearing products (see Appendix I) due to
the absence of default events and losses borne by covered bond investors. This is due
both to the inherent characteristics of covered bonds as a financial product as well as the
support that has been provided at times to their issuers by the Sovereign state backing it.
For example, as the table below indicates, during the crisis, defaults of certain covered
bonds were prevented by the bail-out of their issuers.
Table 1. List of large covered bond issuers subject to a bail-out to safeguard the financial stability
and well-functioning of the covered bond market
Source: EBA
Cover pools in the EU usually consist of residential and commercial real estate
mortgages, and public sector and shipping loans (figure 1). Other types of collateral met
in the market are SME loans, infrastructure loans and aircraft loans. In most Member
States, eligible assets for inclusion in the cover pool are prescribed by legislation and are
usually mortgages and private sector obligations (Table 2).
4
Figure 1. Cover pool composition and new issuance 2003 - 2014
Source: ECBC, Dealogic, own calculations
Table 2. Legal framework for selected European countries regarding cover pool's asset eligibility
Member State Assets eligibility – requirement by law
Denmark
Loans secured by real estate (80% or 75% max LTV); exposures to public
authorities (SDROs also include exposures to credit institutions (15%) and
collateral in ships).
France
OF -First-rank residential and commercial mortgages (max LTV 80%);
state and third party guaranteed real estate loans; public sector exposures;
securitization of the above. OH – Residential mortgages and securitization
of them.
Germany
Mortgages, public sector loans, ship, aircraft (max LTV 60% for all).
Credit institutions exposures (max 10% nominal value of the bond),
derivatives (max 12% of cover assets).
Norway Residential (max LTV 75%) and commercial mortgages (max LTV 60%);
public sector loans; derivative agreements; substitute assets.
Spain Cédulas Hipotecarias(CH): secured by the entire mortgage loan book
(excluding securitizations or loans securing mortgage bonds).
Sweden Residential (max LTV 75%) and commercial (max LTV 60%, max 10% of
cover pool) mortgages; public sector assets; substitute assets.
Source: ECBC, Banco de España (data as of Jan. 2013)
Covered bonds are generally medium-term financial instruments. Currently, the average
maturity of new issues is around 5-7 years with 70% of all new issues maturing within 7
years. The inception of the financial crisis significantly shortened maturities but this was
a one-off market reaction; in 2008 more than 50% of all new issuance was maturing
between 1 and 3 years - almost twice as much than the year before or the year after
(figure 2). Recent observations for 2015 show a renewed tendency towards longer
maturities, justified by the search for higher yields.
5
Figure 2. Maturity profile of CB issuance in the EU (public and private placement)
Source: Dealogic
Almost all EU Member States have active covered bond markets1. Nevertheless, covered
bond issuance is dominated by a few: approximately 80% of global covered bond
issuance is represented by six EU Member States alone (Denmark, France, Germany,
Italy, Spain and Sweden).
Issuers of covered bonds are mainly credit institutions subject to prudential oversight.
Public placements account for about half of the total covered bond market (figure 3).
Fixed coupon covered bonds make up for more than 75% of the total new issuance
(figure 4).
1 Other than Croatia and Estonia
6
Figure 3. Proportion of public versus private
issuance
Source: ECBC statistics. Data Jan 2015
Figure 4. Breakdown between fixed and floating
coupon for new issuance (public placements only)
Source: 2014 ECBC Statistics, Dealogic
There is no specific legislation at EU level which governs all specificities regarding
covered bond issuance2. As Table 3 shows, specific domestic legislation on issuance
usually exists but varies among Member States.
Table 3: Legal framework on Issuance for selected EU Member States
Member State Issuing entity Issuance
limit Over-collateralization limit
Denmark Commercial (SDOs) and mortgage
banks (SDROs) none
108% minimum coverage ratio
(mandatory for mortgage banks and
not for commercial banks)
France
Specialist credit institution -
Sociètes de crédit foncier (SCF)
and Sociètes de Financement de
l’Habitat (SFH)
none 102% minimum coverage ratio
Germany Universal credit institution with an
special license none 102% minimum coverage ratio
Spain
Credit institutions entitled to
participate in the mortgage market
(traditionally commercial,
cooperative and savings banks)
none
125% for Cédulas Hipotecarias
(143% for Cédulas Territoriales (CH
whose cover assets are public
administrations exposures)
Sweden Credit institutions with a special
license none 102% minimum coverage ratio
Source: ECBC, Banco de España (data as of Jan. 2013)
Moreover, such legislation applies mainly to public placements. In contrast, issuers use
private placements in order to include non-eligible assets under specific national
legislations. One recent example is the small and medium-sized enterprises (SMEs)
structured covered bond issued by Commerzbank in February 2013, which replicates
exactly the structure of German legislative covered bonds framework with loans to SMEs
used as collateral assets (non-eligible under German covered bond legislation).
Although there is no dedicated legislation at EU level which governs the operation of
covered bonds, there are various EU provisions affecting investment in this asset class.
More specifically:
2 Covered bond trading is subject to the normal MiFID II rules
49%
1%
44%
6%
Public
Syndicated onlyListedonly
7
The Undertakings for Collective Investment in Transferable Securities (UCITS
Directive), which sets limits and provides exceptions for the assets in which a
UCITS can invest.
The Capital Requirement Directive (CRD IV) which, inter alia, modifies the risk-
weighting approach for covered bonds
The Solvency II Directive which addresses many different sources of risks that
interact with each other and allow for the calculation of a solvency capital
requirement.
The Bank Recovery and Resolution Directive (BRRD) which, in principle,
exempts covered bonds from being written down following a bail-in intervention
of the national authorities and provides that the assets of the cover pool must
remain segregated and well-funded.
A Delegated Act on the liquidity coverage ratio (LCR) adopted by the
Commission on 10 October 2014, under which covered bonds are classified as
one of the relevant liquid assets.
The typology of covered bond investors is diverse. Main investors are credit institutions,
investment funds, pension funds, insurance companies and central banks. In particular,
credit institutions and investment funds account for more than 70% of total market.
Insurance companies and pension funds' investment in this market accounts roughly for
10%.
Figure 5. Typology and share of covered bonds
investors in the EU by financial institution
Source: Natixis, own calculations
Figure 6. Geographic distribution of covered bond
investors – aggregate figures
Source: Barclays Research
In terms of investors' geographical distribution, large cross-border investment and
financial flows have characterised the covered bond market since the development of the
single market and the introduction of the euro. However, the number of Member States
that accounts for the holding of the majority of the total outstanding cover bonds is small
(figure 6).
For an investor's perspective, the benefits from investing in covered bonds are mainly the
dual recourse to the issuer and the cover pool, the covered bonds higher ratings which
8
also remain over time more robust than these of other financial instruments and their
favourable EU legislative treatment when it comes to assessing a financial institution's
liquidity.
Table 4. Pros & cons of covered bonds versus senior unsecured debt from an investor's perspective
On the downside is that covered bonds investors usually have to accept lower yields
when buying covered bonds relative to senior unsecured debt (figure 7). And although
this price difference was even more apparent during the midst of the financial crisis, the
observed spread compression in the market could, among other things, also imply an
increasing risk appetite of market participants.
Figure 7: Asset Swap Margin Difference between iBoxx EUR Bank Senior and iBoxx EUR Covered
Source: Thomson Reuters Eikon
Covered bonds issuance throughout the crisis remained remarkably resilient (figures 1
and 8a) in comparison to other collateralised debt securities in Europe, whose issuance
fell more than 75% since 2008 (figure 8b). At the same time, banks were able to continue
diversifying their funding bases through senior unsecured debt while their funding needs
decreased as they have been deleveraging and/or restructuring. During this period, the
collateral backing of new covered bonds issuance was shifted largely to mortgages. The
significant decline in 2013 in both issuance and outstanding volumes may reflect the on-
going contraction of banks’ balance sheets, the improved availability of alternative
sources of funding, the cancellation of retained deals that had been used as collateral with
central banks and publicly issued deals reaching their maturity with no need for new
issues.
9
Figure 8a. Issuance of private bank debt 2003-2012
Source: Banco de España publications3
Figure 8b. Issuance of other collateralised debt securities4 2003-2014 (€ bn)
Source: European Financial Stability and Integration, April 2015
Since the onset of the sovereign debt crisis, there has been a 'flight to safety', with
German Pfandbriefe seeing higher demand (and thus lower yields) compared to similarly
rated covered bonds in other jurisdictions. Figure 9 shows that until Sep 2012 the 'safety'
of the German Pfandbriefe resulted in lower yields compared with other AAA ranked
covered bonds issued in the euro area. Thereafter the ECB asset purchasing programs
seemed to have smoothed yield differentials even among AAA and AA rating categories.
3 R. Martín, J.Sevillano and L.González, 2013, Banco de España Estabilidad Financiera, No. 24. Only
banks of those countries that have issued at least one covered bond since 2003, according to Dealogic, are
included in the sample. 4 ABS: asset-backed securities; CDO: collaterised debt obligations; MBS: mortgage-backed securities;
CMBS: commercial mortgage-backed securities; RMBS: residential mortgage-backed securities; SME:
small and medium enterprises; WBS: whole business securitisation.
0
100
200
300
400
500
600
700
800
900
2003 2005 2007 2009 2011 2013 2015
10
Figure 9. Covered Bond yields by type of rating (Euro area, 5-year maturity)
Source: Bloomberg
1.1. II. A Benchmark for further EU standardization
Similarities between US and EU frameworks
Government securities play an important role in financial markets. Their attractiveness
for investors lies, among other aspects, on the homogeneity and substitutability between
issues, the normally low credit risk perception compared to private issues (reflecting the
taxation power of governments) and the importance of their yields (as a reference to price
other debt).
To assimilate the low risk status normally conferred to government bonds, non-
government securities have developed characteristics to make them have (or make them
seem as having) lower risk. For instance, they have established (over)collateralization or
other forms of guarantees. In this regard, in Europe covered bonds were developed;
bonds backed indirectly by private mortgages or public sector loans. In the United States,
the government sponsored enterprises5 (GSEs) were established to support and increase
the credit quality of private sector collateral (through implicit or explicit government
guarantees).
To further enhance liquidity, government and non-government issuance is further
standardized / made substitutable by issuing bonds with limited sets of maturities and
relatively large issue sizes (benchmarks). This includes the possibility of re-opening
existing maturities. In addition, quotation can improve transparency and liquidity. In
essence, the measures are meant to achieve the objective of facilitating market
participation and pricing.
5 Student Loan Marketing Association (Sallie Mae), Federal National Mortgage Association (Fannie Mae),
Federal home Loan Mortgage Corporation (Freddie Mac), Federal Home Loan Banks and Federal Farm
Credit Banks.
11
Figure 10. Outstanding liabilities. US GSEs and
Covered Bonds
Figure 11. US Treasury and US GSEs yield curves
Source: U.S. Federal Reserve, European Covered Bond Council Source: Bloomberg
Figures 10 presents the extent to which the alternatives mentioned above have been
successful in developing a market for non-government securities: the total amount of
outstanding (EU) covered bonds and (US) GSE issued bonds. The graph makes evident
the similar size of both markets. It also presents the divergent trend in their respective
evolution as a result of the crisis: US GSE-outstanding liabilities have diminished.
Instead, globally, the volume of outstanding covered bonds, including EU and third
countries, has increased.
Differences
The extent of the success of these markets for non-government securities is not limited to
quantities, however. The impact as a price reference of some of them is probably more
significant. Figure 11 presents the US Treasury (secondary market) yield curve. It also
presents evidence of a yield curve for GSE debt. In particular it presents evidence that
GSE debt:
Covers a large spectrum of maturities (and, consequently, pricing points).
Is comparable, to a large extent, to the breadth of US Treasury debt.
Is resilient (counter-cyclical) under stressed market conditions (for instance, as
experienced in Autumn 2008)6.
The literature has generally recognized several reasons for the breadth and resilience of
GSE debt. Two of them stand out in comparison with the EU covered bond experience:
The standardization of structured debt achieved by the GSEs.
US Treasury implicit or explicit support behind the GSEs.
With regards to the second point, no common EU backstop exists or is expected to exist
to support covered bonds, covered bond frameworks or covered bond issuers7. That is,
the support presented in Table 1, for instance, refers to individual EU Member State
(Treasury) support. Instead, the possibility of further standardizing EU covered bond
frameworks is analyzed below.
6 In particular, on October 2008, considered the peak of the US financial crisis, the spread between 10 year
US Treasury and GSE guaranteed debt stood at 80 basis points. 7 The implicit support was rendered explicit in September 2008.
0
500
1.000
1.500
2.000
2.500
3.000
3.500 US GSEs ($)
Covered Bonds (€)
12
Towards further standardization
Whilst there are five US GSEs, the market is dominated by two of them (Fannie Mae and
Freddie Mac8). These utilize a variety of issuance formats for their securities. Liquidity
of GSEs is enhanced by having established a regular program of debt issuance9, with
broad investor distribution; and having been accompanied by active repo and securities
lending markets for many of these issues. Almost all of the programmed issues are
quoted. Secondary market prices are provided by many domestic and foreign dealers.
However, what is probably most significant of the structured debt market in the United
States is the ability of mortgage lenders across the United States to obtain financing
through primary or secondary purchases of mortgages by the GSEs irrespective of their
place of origin. Conforming loans across the United States that satisfy the conditions
stipulated in GSE asset purchases or securitizations can rely and benefit from the GSE
activities irrespective of the US state where they take place. US GSEs only purchase or
securitize mortgages that "conform" to their established guidelines. Guidelines include
measures of (i) minimum requirements of loan borrowers, including income, debt and
credit score measures; (ii) minimum requirements for loans, including size of the loan,
loan-to-value considerations; etc. Guidelines apply across the US and the single
difference directly considering US geographic differences are considerations driven by
differences in US house prices across counties10
. Collectively these different elements
have allowed the market for GSE debt to flourish and supported the establishment of a
GSE yield curve by markets.
This is far from being the case in Europe and is probably the most important
characteristic of GSE debt vis-à-vis covered bonds in Europe. Covered bond legislation
differs across the continent and the ability for mortgage providers to benefit from
wholesale market funding is very reliant on it. Thus, no similar instrument is hitherto
available in the EU. Covered bonds are a different, wider and altogether much more
heterogeneous asset class than GSE debt. Whilst figure 10 could give the impression that
there is a similar US and European asset class, this is far from being the case. The figure
makes no distinction whether the underlying covered bond debt is considered
substitutable, homogenous, etc. In this regard, figure 12 gives a sense of how covered
bonds are fragmented along observable characteristics, including geographic boundaries,
within and between EU Member States and other third countries.
8 Indeed the term GSE is indistinctively applied to just them two, an aspect which is done in this text too. 9 For instance, most long-term debt is issued in public monthly security sales through designated dealer
groups using both syndicated and auction pricing methodologies. 10 Limits are driven by median home prices estimated by the Federal Housing Administration (FHA) of the
Department of Housing and Urban Development (HUD) and take into account the size of the home.
13
Figure 12. Benchmark issues by country/year (number)
Source: Unicredit research, Association of German Pfandbrief Banks.
Figure 12 also makes evident the extent to which countries issue more or less covered
bonds depending on idiosyncratic characteristics (for instance, their business cycle).
Finally, the lead of a small number of EU Member States (France, Germany and Spain)
in the past, representing most of covered bond issuance, is falling, giving room to more
heterogeneity: they now represent 45% of a much more atomized market (2014 European
Covered Bond Fact Book, with data referred to 2013).
The large differences characterizing covered bonds are further analyzed in the
Consultation Paper and the next sections. The analysis recognizes that differences raised
by the risk exposure that investors face should be priced-in. However, changes driven by
regulatory and legislative differences across the EU that can be ironed out should be
addressed. Whilst there is no inherent conflict in pricing risk according to the geographic
exposure the investors face, the risk should not be driven by the underlying legislation
characterizing and differentiating covered bond frameworks across EU Member States.
1.2. III. Potential Risks and Vulnerabilities
Covered bonds are not a homogenous product. Differences that arise between them can
be driven by several factors. For instance, the different legislative framework defining
covered bonds in Member States is one of them. Another is the different supervisory
approaches and practices present across the EU. Finally, it is possible that common
elements of the regulatory and supervisory frameworks present in the EU interact with
Member States' idiosyncratic characteristics and salient features to generate different
outcomes. This section presents these factors in the context of the risks and
vulnerabilities characterizing covered bonds. In particular, it focuses on three of them:
market fragmentation; transparency and barriers worries to investor entry and
participation; and asset encumbrance.
14
3.1. Market fragmentation
Since the Autumn of 2007, covered bond markets experienced an increased dispersion of
yields. Figure 13 presents Covered bond pricing references available for a number of
Member States whose currency is the Euro. Figure 14 presents the same price reference
with respect to Sovereign yields. The comparison between both figures makes evident a
number of similarities between the behavior of the underlying Sovereign and the covered
bond issued in its jurisdiction.
Figure 13. Covered bond yields Figure 14. Sovereign bond yields
Source: Bloomberg, European Commission services
Prior to 2007, there was a yield contraction between the various European covered bonds
indicating that investors viewed those as fundamentally homogeneous assets, hence of
very similar risk characteristics and high credit quality regardless of the Member State of
issuance. Country factors do not appear to have played a decisive role in investors'
decisions, probably on the assumption that covered bonds were backed by substantially
similar legal frameworks and, in any event, assisted by the strength of European
sovereigns most of which were very highly rated at the time.
The dynamics that existed in covered bond markets changed completely after 2007 by the
significant yield dispersion between the financial instruments of various Member States.
European covered bond markets fragmented along jurisdictional lines and between
stronger and weaker Member States, in the sense that secondary market pricing became
dominated by country factors and favoured covered bonds issued from Member States
viewed as safer jurisdictions.
There are several ways in which perceived EU Member States differences, directly or
indirectly related to covered bond frameworks, can interact and impinge into covered
bond pricing and subsequently cause such yield behaviour.
Hypothesis 1: Price differences due to different fiscal positions of Member States
Covered bonds are asset-backed financial instruments which include a preferential claim
against a dedicated pool of collateral including overcollateralization (the Covered bond's
cover pool). In fact, issuers must ensure that the pool consistently backs the covered bond
and, in the event of default, the investor has recourse to both the cover pool and the
issuer. This is known as the double recourse of covered bonds. Such features can interact
with the fiscal position of Member States in the following sense. The positive track
record behind covered bonds means that public authorities have a significant incentive to
support them and implicitly stand behind them: the fact that no covered bond has ever
defaulted is an obvious quality mark with respect to other financial assets. In fact, the
15
willingness by authorities to support covered bonds from defaulting has come to be
known as the triple recourse of covered bonds.
Unfortunately, public authorities need to support covered bond frameworks via
supporting the credit institutions which have issued covered bonds. Still, the ability to
provide such support (to support the positive covered bond trait of never having
defaulted) differs across Member States. Countries with strong fiscal positions can back
their banking sectors to support covered bond issuers without necessarily exposing their
fiscal position. However, this is not true for all Member States. Instead, so-called stressed
Member States are, for a variety of reasons, more vulnerable in case of extending such
level and degree of support, either because of the state of their public finances or the size
of the support the banking sector would require.
Therefore, if we assume that a single definition for a covered bond were to exist and that
the legal frameworks governing covered bond markets across Member States were the
same, then a reason behind the observed price differences of Figure 13 could be the fiscal
position of a Member State and/or the redenomination risk of a Member State especially
during the peak of the crisis in 2012.
More specifically, the comparison between Figures 13 and 14, clearly indicate that the
European covered bonds, due to the implicit public support for these markets, became a
proxy for sovereign risk. In other words, one could argue that at a point in time, the
pricing of covered bonds became driven predominantly by the Member States’ financial
strength and their banking sectors as a whole, rather than the intrinsic credit quality of the
assets in the cover pool and the financial soundness of the issuer. Moreover, currency re-
denomination fears at the height of the Euro crisis in mid-2012 seems to have affected
markets in the same way.
An additional consideration is that, the interaction between a Member State's fiscal
position and its covered bonds pricing can be almost mechanistic. Notice that figures 13
and 14 differ in one important respect: price changes seem more volatile with respect to
covered bonds. This could be due to several aspects. One of them is the fact that
Sovereign bonds are considered a risk free asset by regulators, unrelated to credit rating
evaluations, except under extreme circumstances. Instead, the perceived risk of covered
bonds by regulators (and, therefore, banks) is driven by credit rating agencies' reports. In
this regard, when covered bonds fall under ECAI 2 ratings, the sovereign rating becomes
a key defining trait of the covered bonds, as credit agencies seem to establish their ratings
by relying on that of the sovereign and not on the individual issuer or issuance rating
(figure 15).
16
Figure 15. Correlation between Sovereign and Covered bonds' yields
Source: Markit. Citi Research, 1-y correlation, 2014
For all these reasons it is believed that fiscal factors have impinged on covered bond
pricing and have ultimately lead to introducing a distortion in covered bond markets.
Hypothesis 2: Price differences due to differences in covered bond regulatory and/or
supervisory frameworks.
i. Differences in regulatory frameworks
The observed price differences among Member States may have also been generated by
the lack of a widely-accepted definition for covered bonds and /or the lack of a similar or
comparable regulatory framework governing their key characteristics. Tables 5 and 6
present evidence of significant regulatory differences in covered bond frameworks and
treatment between Member States.
For example, there are significant differences in the legislative treatment of the legal
segregation of the cover assets from the issuer in the event of insolvency or resolution.
Although such circumstance remains untested, given that no issuers were allowed to fail,
it is possible that it creates doubts among investors in the effectiveness of the dual
recourse mechanism and it reinforces the expectation from the market that covered bond
issuers will be bailed-out.
Other examples of significantly different legislative treatment among Member States can
be observed in the requirements regarding the eligibility of cover assets, the type of
reporting to the supervisor, the national transparency requirements, the limitation on
covered bond issuance, etc. (Schedule II, III and Table 6).
17
Table 5. Legal and Supervisory frameworks in main European Countries
Source: AFI, FUNCAS Spanish Economic and Financial Outlook 2012
Table 6. Limitations to Covered bond issuance from a global perspective
Source: ECBC 2014 Factbook
18
i. Differences in supervisory frameworks
Price differences between Member States may have also been generated because of
differences in Member States' supervisory approaches to covered bonds. Table 5 and
Schedule III show for instance that some competent authorities monitor directly the cover
pool, whilst others monitor the issuing bank, while others exercise an altogether different
approach to monitoring. Moreover, there are also competent authorities that exercise no
monitoring. Hence, differences in the supervisory approach of cover pools between
Member States may create an unintentional market distortion which is ultimately
reflected on covered bond prices. This is besides the possibility that different supervisory
practices, including perceived qualitative practices differences, could also generate
distortions.
To the extent that the crisis revealed institutional weaknesses (for instance in banking
supervision), one could argue, that investors have become unwilling to rely, rightly or
wrongly, on the institutional and legal setting that through special public supervision and
asset segregation was to deliver effective protection to bondholders. They may also have
had – justified or unjustified – doubts about the quality of the assets that credit
institutions held in the cover pools, in particular in Member States undergoing mortgage
market downturns and insufficient transparency on those cover pools would have
exacerbated loss of confidence and the tendency to rely on implicit public support as
described before.
Thus one could argue that a set of common high quality standards for all covered bonds
in the EU, provided that such standards sought to enforce market discipline, could have
been proved useful in mitigating the need of investors to rely on public support and could
protect the covered bond brand.
The need for a minimum common framework that protects the covered bond instrument
has been pointed out previously and, as a result, has led to market initiatives to try to
remedy such situation. In particular, the European Covered Bonds Council (ECBC)
established recently the Covered Bond Label Convention, with core characteristics
required for a covered bond programme to qualify for its quality Label. As of end 2014,
thirteen countries (Austria, Denmark, Finland, France, Germany, Ireland, Italy,
Netherlands, Norway, Portugal, Spain, Sweden and United Kingdom) have implemented
this initiative which covers a market share of approximately more than 50% of the total
covered bonds outstanding.
However, the German Pfandbrief is not part of the initiative. Moreover, it is worth noting
that ECBC's coverage can only go as far as 70% of the total market since for covered
bonds to qualify for the Label they need to be eligible under either the CRD or UCITS
frameworks.
3.2 Transparency and barriers to investor entry
Hypothesis: differences in covered bond regulatory frameworks and legislation limit
information and transparency.
Rules applying to covered bonds vary significantly across Member States. As a result,
this generates a need for investors to undertake separate analyses for each country and for
each structure to understand its unique features. Thus, from an investor perspective,
19
barriers to entry are likely created by the lack of a harmonised regulatory regime as well
as the lack of national transparency requirements. Thus, while domestic investors in
covered bond primary markets are important across most EU jurisdictions, only a limited
number of investors from other Member States (German, French) are active cross-border,
as figure 16 shows.
In this regard, ICMA Covered Bonds Investor Council (CBIC) has raised concerns about
the issue of limited transparency across the EU covered bond markets. The minima
transparency requirements established as a response have been seen as a starting point to
address such concerns, but the CBIC has called for more standardisation of reporting at
European level to help investors compare information. This includes investors from third
countries, as the absence of a concrete framework to facilitate comparison and with
which to establish equivalence assessments with third country covered bonds regimes has
also restricted cross-border investment flows. As a consequence, third country
investment in the European covered bond markets remains limited (see figures 16, 17 and
Schedule IV).
Figure 16: Breakdown of investors by country for each CB market (new benchmark issues)
Source: BofA Merrill Lynch, 2014
Why investors from some specific Member States invest across the EU could be due to
several reasons. One could be that, only economies such as France and Germany, with
large investor bases, can look through the barrage of rules and legislative differences to
invest. Another could be that the development of covered bond frameworks in some
particular economies has made their investor bases particularly welcoming to such
financial products. Investors have become supportive of covered bonds, for the safety
they are expected to represent and have led such demand. Differences in legislation are
perceived as a nuisance but considered minor in comparison to the benefits provided by
covered bonds across time, including the expectation that Member States' public
authorities will do their outmost to protect the covered bond label. In this regard, a move
towards further standardisation and harmonization would benefit investors but
particularly public authorities to limit the existing wedge present between the benefits
expected from covered bonds (close to being a risk free asset) and the (implicit) need by
public authorities to support them.
Relatedly, in the context of the financial crisis, the benefits derived from the security
provided by covered bonds have justified the proliferation of covered bond frameworks
beyond the EU to markets which did not have CB legislation or were banks did not
traditionally rely on covered bonds to obtain financing. This evolution can be positive, in
a context where cooperation and understanding between different authorities does not
20
undermine the covered bond label. However, there is a parallel evolution which could
challenge the covered bond label and its perceived benefits.
Instances of new financial assets marketed as covered bonds have developed in some
jurisdictions. Financial innovation may blur the lines between what are covered bonds
and look-a-like covered bonds. In this regard, they could very well eventually undermine
investors' confidence in covered bonds as an asset class. For instance, some programmes
for structured "covered bonds" have been developed, in which a pool of SME financing
provides collateral for one bond. The protection provided by the collateral is not
legislative-based, but arising from a pure contractual arrangement made by the issuer.
Whilst such evidence is small and scattered to represent a challenge, it is nevertheless
revealing. Such could very well challenge covered bond markets. To this end there may
be merits in harmonising national transparency requirements for investors to be fully
aware of the product they are purchasing.
Figure 17. Allocation of total EUR Covered bonds
Source: ECBC 2014 Factbook
Finally, there could be benefits from further transparency, including to public authorities.
Well-developed and functioning financial markets are characterised by the capacity of
investors to access information to understand the exposure they take if they purchase a
financial asset. But transparency can also help and benefit public decision-making.
During the financial crisis the ability to issue covered bonds in primary markets by listing
them and gathering a sufficient number of banks to syndicate an issue become more
restricted. Hence the remedy known as private placements, whereby a limited number of
investors purchase a covered bond not necessarily intended to be actively traded.
However issuers also retained covered bonds for themselves, with the objective of using
them to obtain financing from central banks as collateral, instead offinding willing
investors to purchase them.
21
Figure 18. Outstanding volume of GBP denominated
covered bonds over time in GBP bn
Figure 19. Outstanding volume of GBP denominated
covered bonds over time in EUR bn
Source: ECBC Source: ECBC, Banks' annual reports
The above practice obviously affects the liquidity and transparency in financial markets.
Figures 18 and 19 present the level of public versus private placement depending on
whether one relies on banks' annual reporting (figure 19) or analysis undertaken by
analysts (figure 18). The differences are, obviously, striking. And the lack of
transparency can work to the detriment of private and public interests. For investors
benefit from having a clear idea of what and where to invest. But public authorities also
benefit from understanding how and to whom are covered bonds being issued,
particularly if there is an explicit third recourse.
In general, encouraging greater transparency and disclosure could be seen as a first and
cautious step to improve covered bond market functioning in Europe.
3.3 Asset encumbrance
Any bank's issuance is heavily dependent on the underlying legislative framework
defining the liability. However, some bank business models have been specifically
conceived to rely almost exclusively on covered bonds; for covered bond issuers include
what are known as universal banks, but also specialized mortgage banks and special
purpose vehicles. In this regard, the heavy reliance in some Member States on specific
business models has been seen as a concern. For reliance on covered bonds can trigger
significant amounts of what is known as asset encumbrance.
Covered bonds play an important role in supporting domestic financial stability: issuers
retain the credit risk of underlying loans and, hence, issuance disincentives moral hazard.
Investors benefit from establishing a cover pool of assets. However, such pools of assets,
also known as "encumbered", can come at the expense of issuers (and public authorities
backing them). And, during the crisis, covered bond issuance exceeded senior unsecured
issuance in the EUR markets for the first time in history. As a result, asset encumbrance
became a concern: a large amount of bank assets, pledged to support specific creditors,
would not be available in insolvency, making banks more vulnerable. This was
particularly relevant in Europe, given the preponderance of banks in its financial system.
Member States' supervision on asset encumbrance or new covered bond issuance levels is
primarily designed and focused on domestic financial stability. On the other hand,
Directive 2014/59 (Bank Recovery and Resolution Directive) can act as an indirect limit
on asset encumbrance11
. Still, the differences in the framework of new covered bond
issuance levels among Member States (see Table 6) and the cumulative importance of
11 Directive 2014/59 (the "BRRD") requires that all credit institutions, including covered bond issuers,
meet a minimum requirement in terms of liabilities available for bail-in.
22
these markets for financial stability beyond the domestic borders probably need to be
thoroughly addressed.
Figure 20. Asset encumbrance (evolution between 2007 and 2013 per national banking sector and
expressed as a percentage of total assets)
Source: IMF Global Financial Stability Report October 2013
Figure 20 presents the increase in asset encumbrance that has taken place due to several
bank funding patterns. Asset encumbrance through covered bonds has been increasing
and is particularly high in Denmark, reaching 35% of total assets of DK banks12
. Levels
of asset encumbrance could rise further if a broader scope of assets were allowed in the
cover pools, a trend largely resisted so far by covered bond laws but which could take off
in the market with structured products using the "covered bond label" on the back of
currently ineligible asset classes. Moreover, whilst concerns related to asset encumbrance
are related to transparency concerns and the ability to understand bank balance sheets and
exposures, they are a separate issue: transparency encompasses a wider set of points, but
it cannot address asset encumbrance fears by itself.
1.3. IV. Summary
Covered bonds were far less affected than other financial markets by the crisis and
remained for the most part a resilient and reliable source of funding for credit institutions
throughout significant stressed conditions. Covered bonds performed better in this
respect relative to unsecured debt and also to other forms of collateralized lending.
Key to such good performance is the perception that covered bonds are very low risk
instruments, which is indeed backed by the absence of credit losses for investors
12 Danish mortgage credit institutions, some of which are classified as systemically important financial
institutions, do not take deposits and cannot access money markets. They only grant mortgage loans,
which are funded through the issuance of covered bonds.
23
throughout the whole history of this instrument although the bail-out of certain issuers in
financial distress undoubtedly contributed to preserving such unblemished record.
The undisputed strengths of European covered bonds should not, however, mask certain
underlying vulnerabilities and challenges:
a) Market fragmentation: Prior to 2007, yield contraction between the various
European covered bonds showed that investors viewed those as fundamentally
homogeneous assets, hence of very similar risk characteristics and high credit
quality regardless of the Member State of issuance. However, the dynamics that
existed in covered bond markets changed completely after 2007 by the significant
yield dispersion between the financial instruments of various Member States.
Arguably, European covered bond markets fragmented along jurisdictional lines
and between stronger and weaker Member States, in the sense that secondary
market pricing became dominated by country factors and favoured covered bonds
issued from Member States viewed as safer jurisdictions, regardless of the actual
credit quality of cover assets or the financial strength of individual issuers.
b) There is insufficient homogeneity in legal and supervisory frameworks which
forces investors to incur in higher costs to undertake separate analysis for the
covered bonds of each Member State and which may partly explain why the
covered bond investor base remains relatively home-biased and concentrated in a
few large Member States.
c) The lack of a truly integrated European covered bond model may also be
hampering investment from third countries, as investors there do not have a
comprehensive basis for comparison with the covered bond framework of their
home jurisdiction. Third country investors may be particularly prone to
indiscriminate retrenchment in the event of suffering losses, with the risk that
losses in relation to the covered bonds of a Member State may become a stigma
associated to the covered bonds of other or all Member States.
d) Reporting to investors is subject to inconsistent transparency requirements, not
only among Member States but also asset classes and the overall levels of
disclosure in relation to the cover pool are deemed in general very poor by many
market commentators when compared to the disclosures required for other
structured financial instruments (e.g. ABS). As a matter of fact, efforts to enhance
transparency and standardisation in covered bonds may encounter difficulties in
the future resulting from the market trend towards increased private placements
and tailor-made structures noted above.
Finally, the benefits of covered bonds as a source of funding have the flip side of
encouraging asset encumbrance in credit institutions' balance sheets. Asset encumbrance
raises concerns among supervisors and policy makers for two main reasons:
it leads to the unavailability of assets to support the resolution of credit institutions
and has the potential to inflict costs on unsecured creditors and taxpayers; and
it reduces the pool of assets available to the issuer to obtain liquidity in the event of
unforeseen stresses.
Levels of asset encumbrance could rise further if a broader scope of assets were allowed
in the cover pools, a trend largely resisted so far by covered bond laws but which could
24
take off in the market with structured products using the "covered bond label" on the
back of currently ineligible asset classes. Still, covered bonds are only one source of asset
encumbrance, which needs to be considered holistically taking into account other forms
of asset-backed and secured funding and taking into account the relative importance of
unsecured funding of any given credit institution.
25
1.4. V. Schedules
SCHEDULE I – COVERED BONDS ARE LOW RISK YIELD-BEARING INSTRUMENTS
Annualised volatility by sector: 2010 -2014 (april)
Source: BoFA Merrill Lynch Global research
SCHEDULE II –LEGAL FRAMEWORK FOR SELECTED EU MEMBER STATES REGARDING ASSET
ELIGIBILITY (AS OF JANUARY 2013).
Member State Assets eligibility – requirement by law
Denmark
Loans secured by real estate (80% or 75% max LTV); exposures to public
authorities (SDROs also include exposures to credit institutions (15%) and
collateral in ships).
France
OF -First-rank residential and commercial mortgages (max LTV 80%);
state and third party guaranteed real estate loans; public sector exposures;
securitization of the above. OH – Residential mortgages and securitization
of them.
Germany
Mortgages, public sector loans, ship, aircraft (max LTV 60% for all).
Credit institutions exposures (max 10% nominal value of the bond),
derivatives (max 12% of cover assets).
Norway Residential (max LTV 75%) and commercial mortgages (max LTV 60%);
public sector loans; derivative agreements; substitute assets.
Spain Cédulas Hipotecarias(CH): secured by the entire mortgage loan book
(excluding securitizations or loans securing mortgage bonds).
Sweden Residential (max LTV 75%) and commercial (max LTV 60%, max 10% of
cover pool) mortgages; public sector assets; substitute assets.
Source: ECBC, Banco de España
26
SCHEDULE III – SUPERVISORY PRACTICES IN SELECTED MS
Member
State Supervisor
Supervisor's involvement in the
monitoring of the cover pool
Type of reporting to the supervisor /
national transparency requirements
Denmark
Danish Financial
Supervisory
Authority (FSA)
The issuer monitors the cover pool
continuously. Mortgage banks – internal
auditor. Commercial banks - report directly
and quarterly to the FSA (verified by an
external auditor). Issuers must prepare
quarterly reports on asset-liability
management for the FSA.
Investor reports, trading venues and
investor relations websites. National
transparency template (complements the
ECBC label).
France
French Banking
supervisory
authority
Specific controller appointed by the issuer
and agreed by the supervisory authority.
Duties: control eligibility, composition and
valuation of assets; compliance with
minimum coverage ratio (quarterly report);
control management of risks on assets
(liquidity, interest rate, currency and
maturity mismatch risks).
Issuer should issue periodic financial
information (quarterly asset report,
semi-annual report on coverage ratio and
other legislative limits, annual reports on
assets and methods of valuation and a
report on risk management).
Germany
BaFin, German
Federal Financial
Supervisory
Authority
Certified auditor appointed by BaFin. BaFin
must monitor the cover pool on average
every 2 years. Pfandbierdbanks must carry
weekly over collateralization stress tests &
daily calculation of 180-day.
Legal requirement: Quarterly disclosure
of CBs outstanding and characteristics.
Information about their legal structure.
Spain Bank of Spain Issuer must monitor the cover pool (as part
of its risk management and auditing).
Monthly CB report to the Bank of Spain.
Annual accounts contain the details of
the register of loans. National
transparency template consistent with
ECBC label.
Sweden
Swedish
Financial
Supervisory
Authority
(SFSA)
Independent cover pool trustee (appointed
by the SFSA). Duties: monitor the register
and compliance with market and matching
risks. It must submit an annual report to the
SFSA.
Quarterly information about the cover
pool and outstanding CBs (issuer
website). Steps toward national
transparency template.
Source: ECBC, Banco de España (data as of Jan. 2013)
27
SCHEDULE IV - Geographic distribution of covered bond (CB) investors in selected
European markets
Figure I: Geographic distributors of CB investors – France Figure II: Geographic distributors of CB investors –
Germany
Figure III: Geographic distribution of CB investors – Ireland Figure IV: Geographic distribution of CB investors
- Italy
Figure V: Geographic distribution of CB investors – Portugal Figure VI: Geographic distribution of CB investors –
Spain
Source: Natixis